In recent years, aesthetics – like many other philosophical areas – has gradually replaced conceptual analysis projects with theory construction projects. For example, in a presidential speech of the American Society for Aesthetics, Kendall Walton advocates for the theory-construction methodology, which does not primarily aim to capture the meaning of aesthetic terms in ordinary English.Footnote 1 Instead of trying to define what beauty or art is, philosophers have shifted their focus to explaining aesthetic phenomena that arise from our interactions with narratives and artworks.Footnote 2 We are experiencing a shift from what Jonathan Weinberg and Aaron Meskin call the ‘traditional paradox-and-analysis model’ to a new paradigm, the ‘phenomenon-and-explanation model’.Footnote 3 The methodology of the new paradigm explicitly takes its cue from the sciences: look for observable data, propose theories that aim to explain the data, adjudicate competing theories, and repeat.
Despite this shift, there is surprisingly little work on aesthetic explanations. Perhaps the current dearth of writings on this topic can be attributed to the dominance of the traditional paradox-and-analysis model in aesthetics. However, as the phenomenon-and-explanation model becomes more prominent in aesthetics, it becomes increasingly important to investigate the nature of aesthetic explanations.
Given that the methodology of the phenomenon-and-explanation model explicitly takes its cue from the sciences, this paper starts the investigation by looking to recent developments in philosophy of science. In recent decades, philosophers of science are increasingly turning away from conceptions of laws and explanations that are devised in metaphysicians' armchairs to conceptions of laws and explanations that are developed with attention to actual scientific practices, especially practices in the special sciences. One prominent picture that emerged is a pragmatist and pluralist view of scientific explanations.
Taking the methodological similarities between the new paradigm of aesthetics and the sciences as the starting point, I advocate a pragmatist and pluralist view of aesthetic explanations. To bring concreteness to this discussion, I focus on the case of genre explanations: explanations of aesthetic phenomena that centrally cite a work's genre classification. Even though some philosophers have given genre explanations of aesthetic phenomena, others have categorically dismissed genre explanations, calling them unhelpful at best and meaningless at worst. Of the opponents of genre explanations, Gregory Currie most clearly states the theoretical grounds for categorically dismissing genre explanations.Footnote 4 However, I argue that these theoretical grounds do not stand up to scrutiny once we incorporate the central insights from recent works on scientific explanations. On a pragmatist and pluralist view of aesthetic explanations, there is room for genre explanations. In fact, the reasons for accepting genre explanations alongside other kinds of aesthetic explanations are also reasons for accepting a pragmatist and pluralist view of aesthetic explanations.
§1 introduces genre explanations and Currie's arguments against them. §2 draws on one aspect of Currie's arguments to develop the robustness challenge for genre explanations: given that the generalizations underlying genre explanations appear to admit of many exceptions, they seem incapable of supporting as wide a range of counterfactuals as law-like generalizations in other domains can. §3 looks to recent works on ceteris paribus laws and counterfactual robustness to respond to the robustness challenge. §4 draws on another aspect of Currie's arguments to develop the informativeness challenge for genre explanations: given that genre classifications are metaphysically grounded in lower-level features of works, genre explanations seem incapable of being more informative than, or even as informative as, the aesthetic explanations that only cite those lower-level features. §5 looks to recent works on levels of explanation to respond to the informativeness challenge. To conclude, §6 uses the case study of genre explanations to say what it means to be a pragmatist and a pluralist about aesthetic explanations.
1. Genre Explanations
As introduced earlier, genre explanations are explanations of aesthetic phenomena that centrally cite a work's genre classification. In this section, I give further characterizations of genre explanations as a kind, first with simple but illustrative examples and second with references to genre explanations that have been given for specific aesthetic phenomena. I then review Currie's arguments against genre explanations as a kind.
1.1. Lab Specimens
Consider first a case in which the genre of a dance performance explains the appropriate audience response. Suppose you are seeing a dance performance, performed by ordinary people wearing ordinary clothes and doing ordinary things like walking, standing, and sitting down. That's it.
If this is a performance of modern dance, then the appropriate response is probably a mishmash of boredom, confusion, and perhaps even annoyance. There is no recognizable technique. There is no awareness of rhythm and tempo. There is no narrative, and not even any movement that can arouse some emotions. Everything is so ordinary.
But if this is a performance of postmodern dance – indeed, I am (rather minimally) describing Steve Paxton's Satisfyin' Lover (1967) – then the appropriate response is quite different. How curious! How interesting! As a member of the Judson Dance Theater, Paxton and his contemporaries – Yvonne Rainer, Elaine Summers, etc. – consciously rejected the aesthetics of modern dance and sought to legitimize ordinary movements as dance. Ordinary movements, even when performed by ordinary people, can have aesthetic interest for the viewers – in itself, without embodying any narrative or overtly arousing any emotion. Far from boredom or annoyance, the appropriate response to this performance includes puzzled curiosity and cognitive interest.
A promising explanation of the appropriate response to Paxton's Satisfyin' Lover appeals to its classification as a postmodern dance performance. Specifically, this explanation appeals to the generalization ordinary movements in postmodern dance warrant puzzled curiosity and cognitive interest. There may be more to completely explaining the appropriate audience response to this piece. However, for my purpose, it is enough that genre partially explains the appropriate response to this dance performance.
Consider next a case in which the genre of a film explains the appropriate audience response. Suppose you are watching a gory scene in a film, in which a character is being decapitated in gruesome details.
If this scene is a part of a straight-up horror film, then the appropriate response is to scream. Horror films elicit fear in their audiences. As part of a horror film, a gory decapitation scene is scream-worthy. But if this scene is a part of a horror comedy film, then an appropriate response is to laugh. Horror comedies elicit amusement in their audiences. As part of a horror comedy, a gory decapitation scene is laughter-worthy.Footnote 5
Again, a promising explanation of the appropriate response to a gory decapitation scene in, say, Evil Dead 2 appeals to its classification as a horror comedy. Specifically, this explanation appeals to the generalization decapitation scenes in horror comedies warrant laughter.
1.2. Wild Beasts
More complicated genre explanations can be found in the philosophical aesthetics corpus. Philosophers have appealed to genre to explain the following aesthetic phenomena: comedic force of jokes Footnote 6, ethical criticism of art Footnote 7, imaginative resistance Footnote 8, criticism and evaluation Footnote 9, perceived realism in fictions Footnote 10, and moral persuasion Footnote 11. Moreover, style – a close relative, if not a subset, of genre – is also invoked in explaining interpretation of pictorial art Footnote 12 and understanding of theatrical performance Footnote 13. In this paper, I will use the two simple examples of genre explanations provided earlier to indicate and illustrate what can be said about these more complicated genre explanations.
A clarification is in order: my use of the term ‘genre’ broadly corresponds to what Kendall Walton calls ‘category of art’.Footnote 14 A genre is a special grouping of fictions that is recognized by a community as such.Footnote 15 On this inclusive definition of ‘genre’, aesthetic explanations that appeal to style, period, etc. all count as genre explanations.
Nothing substantive hangs on this terminological choice. My following defense of genre explanations applies, mutatis mutandis, to any other aesthetic explanation that centrally cites a work's classification, such as its style or its period. Hence, even if one adopts a less inclusive definition of ‘genre’, genre explanations still stand as exemplars of a broader kind of aesthetic explanation.
1.3. Opponents
Since there has been little work so far on the nature of aesthetic explanation as such, genre explanations have been rarely criticized as a kind. Instead, criticisms of genre explanations typically come as arguments against some specific genre explanation of some specific aesthetic phenomena. It is not possible to address such specific criticisms without getting deep into the first-order debate. So, regrettably, I will set them aside.
However, there are theoretical commonalities to the specific criticisms. These theoretical commonalities are most clearly brought out by Gregory Currie, who stands out for focusing on the theoretical grounds for rejecting genre explanations as a kind. Currie argues that we should categorically reject genre explanations because they are neither informative nor robust. I will briefly review Currie's arguments to bring out two theoretical challenges that, in my view, drive the criticisms of genre explanations. I will then focus on these theoretical challenges for the core of the paper.
To begin, Currie says that genre explanations are not as informative as individualistic explanations, or explanations that appeal to metaphysically fundamental features of a specific work:Footnote 16
Suppose we want to explain the effect of the work on the audience, and cite its being a tragedy. The objector will say that what really matters for explaining the effect of the work is the specific way it is (together with, perhaps, similarities between these specific ways and specific ways possessed by other specific works the people in the audience are familiar with). On this view, the work's being a tragedy does not explain anything left unexplained by the individualistic explanation. Indeed, the individualistic explanation explains more; different tragedies affect their audiences in different ways, and the individualistic explanation cites details capable of accounting for these differences. We need not rest content noting that the effect was ‘generally of the kind we expect from a tragedy’.Footnote 17
Furthermore, while some explanations that relatively lack informativeness make up for the vice through other virtues, genre explanations do not. Specifically, Currie says genre explanations also lack the explanatory virtue of robustness: genre explanations are not as robust as sociological explanations:
In interesting cases, explanation by appeal to genre does not provide the sort of information about counterfactual states of affairs that explanation by appeal to industrialization does [‘industrialization’ is short for the generalization the popularity of organized religions declines when the population shifts due to industrialization]. Hamlet has the effects we associate with a Shakespearean tragedy, but its having them is due to highly specific and contingent features that its being a tragedy tells us nothing about; it could fail utterly to have these effects and still be a tragedy. While industrialization is counterfactually robust, being a tragedy is counterfactually fragile, or relatively so. …Why is Hamlet so intellectually and emotionally affecting? An informative answer may cite the fact that it is a tragedy, but no informative answer will be robust under changes to any of a vast range of details about the play: a small word change here or there would have altered the effect significantly.Footnote 18
Genre explanations fail to be robust because, not only do they admit of some exceptions, they are in fact shot through with numerous exceptions. Given their relative lack of informativeness and robustness, Currie concludes that genre explanations ought to be of no interest to aestheticians.
2. The Robustness Challenge
We start with the theoretical criticism that genre explanations, as a kind, are relatively fragile. We can illustrate this criticism by returning to the two simple examples. Aren't there many boring and annoying postmodern dance pieces? Indeed, in Claudia La Rocco's review of a recent Judson Dance Theater retrospective, The New York Times's dance critic knocked Carolee Schneemann's Lateral Splay (1963) as ‘an amusing but slight exercise in task-based choreography’.Footnote 19 In performances of such pieces, ordinary movements warrant neither puzzled curiosity nor cognitive interest. Similarly, aren't there many unfunny horror comedies, such as all the films in the Scary Movie franchise? In such films, a gory decapitation scene does not warrant laughter. Worryingly, genre explanations appear to lose their explanatory force easily.
The robustness challenge indirectly questions the explanatory worth of genre: it points to the lack of counterfactual robustness to indicate that genre explanations are not genuine explanations. Among the distinctive roles that laws and law-like generalizations perform in science are supporting robust counterfactuals and grounding genuine explanations.Footnote 20 Counterfactual robustness thus goes hand in hand with genuine explanatoriness. Purported explanations that are not counterfactually robust are therefore unlikely to be genuinely explanatory because they are unlikely to be grounded in laws or law-like generalizations.
This challenge to genre explanations is driven by the no-exception intuition: genuine explanations must involve exceptionless laws or law-like generalizations. Since the generalizations that genre explanations cite are apparently shot through with exceptions, they can neither support robust counterfactuals nor ground genuine explanations. As Currie reminds us, just think of cases where two works of the same genre produce significantly different effects on audiences, or cases where small changes to a work make a genre generalization that previously applied to no longer do so. Such cases suggest that genre generalizations are not counterfactually robust. In turn, this lack of counterfactual robustness suggests that genre explanations are not genuinely explanatory. To answer the robustness challenge, we must show that genre generalizations can support robust counterfactuals despite the numerous apparent exceptions to them.
3. Genre Laws and Counterfactual Robustness
The no-exception intuition behind the robustness challenge is misguided. It fails to acknowledge the important role that ceteris paribus laws, or cp-laws, play in actual scientific practices, especially practices in the special sciences. Although cp-laws are apparently shot through with exceptions, they nevertheless support a wide range of counterfactuals. Special scientific cp-laws can support genuine explanations in the special sciences, despite the numerous exceptions that they apparently admit of. I argue that genre laws are best understood as cp-laws. As is the case with other cp-laws, they are counterfactually robust despite the apparent exceptions that they admit of. In this section, I clarify the nature of cp-laws and then consider how they help in responding to the robustness challenge.
3.1. Ceteris Paribus Laws
We start by seeing what ceteris paribus clauses do. Consider the generalization fish eggs develop into fish. This generalization is apparently shot through with exceptions: some fish eggs get ennucleated with sheep DNA and become sheep, some fish eggs get eaten and become nutrients for a turtle, and some fish eggs get irradiated and turn into strange and dysfunctional piles of flesh.Footnote 21 Yet, despite these apparent exceptions, this generalization seems true. That is because we do not implicitly understand the generalization to be making the evidently-false universal claim that all fish eggs turn into fish. Instead, we tacitly understand the generalization to mean that ceteris paribus, fish eggs turn into fish; the apparent exceptions are not genuine counterexamples to the generalization because they are already excluded by the ceteris paribus clause.Footnote 22
Although it would be practically, if not theoretically, impossible to list every one of the infinite number of trajectories a fish egg might take that the ceteris paribus clause excludes, this generalization has a determinate meaning. Importantly, meanings of generalizations like this do not rest on the statistical typicality of the respective standard cases. The number of fish eggs that do not turn into fish is likely to be greater than the number of fish eggs that do. Instead, meanings of generalizations like this ultimately depend on our tacit understanding of what would constitute genuine counterexamples and what would be mere apparent exceptions – even if such tacit understanding is rarely, if ever, fully articulated. We only understand what the generalization fish eggs develop into fish means because we tacitly understand which trajectories of fish eggs are relevantly like the apparent exceptions listed earlier, and so should be excluded by the ceteris paribus clause.
This reliance on our tacit understanding of the ceteris paribus clause may seem rather unsatisfying at first. Can we not explicate the infinite number of cases that are relevantly like the apparent exceptions listed earlier? Even if we can in theory, as Marc Lange points out, our capacity for making such an explicit list would still be itself derived from our tacit understanding of what would constitute genuine counterexamples to this generalization and what would be mere apparent exceptions:
But in what sense would such an expression really be fully explicit? It would derive its content in just the way that the original qualifier did: by virtue of our implicit background understanding of what would count as compelling reasons for (or against) the correctness of applying it to a given case.Footnote 23
The meaning provided by such an explicit list cannot be anything over and above the meaning provided by our tacit understanding of the ceteris paribus clause because the former is dependent on the latter.
Coming to a shared understanding of a ceteris paribus clause is understandably difficult. We would have to come to agreements on what the canonical examples of exceptions are, how we can compare a novel case to the canonical examples, and whether a given novel case counts as a genuine counterexample to the generalization or a mere apparent exception excluded by the ceteris paribus clause. However, trying to come to agreements on these matters is simply the standard mode of operation in actual scientific practices, especially in the special sciences.Footnote 24 Our tacit understanding of particular ceteris paribus clauses is central to scientific investigations.
Call a non-accidental generalization that contains a ceteris paribus clause a ceteris paribus law, or cp-law for short.Footnote 25 There is a pragmatic dimension to the ceteris paribus clause: it ‘[restricts] the law's application to certain purposes’.Footnote 26 Cp-laws only do the works that laws are thought to do, such as supporting counterfactuals and grounding explanations, in suitable contexts – namely, contexts where the disturbing factors that the ceteris paribus clause excludes are irrelevant. Whether a particular cp-law can ground an explanation partly depends on the question that we are asking, or what we are trying to explain.
The range of counterfactuals that cp-laws support is neither identical to nor narrower than the range of counterfactuals that other laws, such as the fundamental laws of physics, support.Footnote 27 We can see this point via an example.Footnote 28 Suppose that the popularity of organized religions declines when the population shifts due to industrialization is a sociological law. Then it supports the counterfactual if the population were to shift due to industrialization, then the popularity of organized religions would decline. First, there are scenarios where fundamental laws of physics hold but our example sociological law does not. For example, suppose that the fundamental laws of physics are the same as they actually are but that human beings are psychologically incapable of following organized religions. Then population shifts due to industrialization would have no effect on the popularity of organized religions. Second, there are scenarios where fundamental laws of physics do not hold but our example sociological law does. For example, suppose that some fundamental parameter of physics is just slightly different from the way it actually is, but without any downstream effects on human sociological behavior. Then the population shifts due to industrialization would still have the same exact effect on the popularity of organized religions.
Whether our example sociological law is a genuine (cp-)law is an open question, depending on whether it supports a stable range of counterfactuals in conjunction with other sociological laws. The goal here is only to show that cp-laws of autonomous social sciences can support robust counterfactuals that range over distinct sets of scenarios, none of which is wholly contained in any other. Hence, this sociological law cannot be said to be less robust than, say, fundamental physics – contrary to what the no-exception intuition suggests.
3.2. Response to the Robustness Challenge
Now, let us return to the robustness challenge. Counterfactual robustness functions as an indicator of genuine explanations. To answer this challenge, we must show that genre laws can support robust counterfactuals despite the numerous apparent exceptions to them. Conceiving of genre laws as cp-laws allows us to do so.
To develop a strategy for responding to the robustness challenge, we start with the postmodern dance case from §1. Consider the generalization ordinary movements in postmodern dance warrant puzzled curiosity and cognitive interest. As is the case with fish eggs develop into fish, we should not construe this genre generalization as an evidently-false universal claim, that all ordinary movements in postmodern dance warrant puzzled curiosity and cognitive interest. Instead, we should understand this genre generalization to mean that ceteris paribus, ordinary movements in postmodern dance warrant puzzled curiosity and cognitive interest. As explained earlier, this genre generalization is meaningful as long as we have an implicit understanding of what the ceteris paribus clause excludes, even if we can never fully articulate this understanding.
Suppose now that ordinary movements in postmodern dance warrant puzzled curiosity and cognitive interest is a genre law. (Whether it is in fact non-accidentally true is an open question; the point here is only to illustrate the response strategy.) As the litter of unsuccessful postmodern dance pieces shows, an ordinary movement in a postmodern dance piece could fail to warrant puzzled curiosity and cognitive interest due to the presence of a number of disturbing factors: lack of innovation from the choreographer, poor execution by the dancers, etc. This genre law therefore admits of numerous – in fact, an infinite number of – apparent exceptions. Despite apparent exceptions like these, this genre law still holds because our implicit understanding of the ceteris paribus clause allows for the rejection of the numerous apparent exceptions as genuine counterexamples, given the presence of a disturbing factor with each apparent exception.
This genre law is also counterfactually robust. The counterfactual it supports, if an ordinary movement were in a postmodern dance, then it would warrant puzzled curiosity and cognitive interest, ranges over a wide variety of scenarios, including scenarios where fundamental laws of physics fail to hold. For example, suppose that some fundamental parameter of physics is just slightly different from the way it actually is, but without any downstream effects on human aesthetic behavior. Then an ordinary movement in a postmodern dance would still warrant puzzled curiosity and cognitive interest. The range of this genre counterfactual is thus no narrower than the range of counterfactuals associated with the fundamental laws of physics. For analogous reasons, the range of this genre counterfactual is also no narrower than the range of counterfactuals associated with sociological laws, biological laws, or indeed any other special science laws. Therefore, this genre law is no less robust than the laws in other domains.
We now have a strategy for responding to the robustness challenge: understanding the generalizations that genre explanations appeal to as cp-laws. Before moving on, let me emphasize two points concerning the foregoing discussions.
First, I am not assuming that the notion of cp-laws, or indeed the notion of ceteris paribus, is unproblematic. Rather, the centrality of cp-laws to the special sciences shows that whatever problems cp-laws have, they are everyone's problems. The no-exception intuition is misguided because it is insufficiently attentive to actual scientific practices, especially practices in the special sciences. Once we reject the no-exception intuition, as we must, we can see that there is nothing uniquely problematic about the appeal to cp-laws in aesthetic explanations.
Second, I am not claiming that all generalizations that cite genre are genre laws. Rather, whether a generalization that cites genre is a genre law, or even whether there are any genre laws at all, is an open question. We can only answer this question by performing tasks typical of actual scientific practices: coming to an agreement on what the canonical examples of exceptions are, how we can compare a novel case to the canonical examples, and whether a given novel case counts as a genuine counterexample to the generalization or a mere apparent exception excluded by the ceteris paribus clause. In responding to the robustness challenge, I am only explaining why genre explanations, as a kind, cannot be ruled out as good aesthetic explanations simply because they appeal to generalizations that cite genre. The a priori and categorical rejection of genre explanations for their alleged counterfactual fragility is not justified. Ultimately, assessing whether a particular genre explanation counts as a good aesthetic explanation demands that we investigate the phenomenon that it purports to explain. In slogan form: evaluating the worth of a specific genre explanation requires doing aesthetics, not meta-aesthetics.
3.3. Weaker Motivations?
One might wonder whether the robustness challenge can be motivated via weaker, more plausible intuitions.Footnote 29 For one, it could be that while explanations can admit of some exceptions, genre explanations admit of too many exceptions. For another, it could be that, unlike scientific explanations, we have no tacit understanding of which exceptions count as genuine and which exceptions count as merely apparent when it comes to genre explanations.
Take the too-many-exceptions thought first. The difficulty with this thought is that it demands a quantitative comparison of exceptions. However, given combinatorialism about possibilities, the number of exceptions is typically either zero or (countably) infinite. For example, in the postmodern dance case, the ceteris paribus clause excludes an infinite number of ways that an ordinary movement in a postmodern dance can fail to warrant puzzled curiosity and cognitive interest due to the presence of a disturbing factor. For another example, in the fish egg case, the ceteris paribus clause excludes an infinite number of trajectories that a fish egg might take which do not result in a fish due to the presence of a disturbing factor. Since exceptions multiply infinitely, once a ceteris paribus clause excludes one apparent exception, it typically excludes an infinite number of them. Therefore, it is difficult to make sense of the not-too-many-exceptions complaint without reducing it to the no-exceptions complaint.
Take the no-tacit-understanding thought second. Let us grant that, with some purported genre explanations, we may not have an adequate tacit understanding of which exceptions count as genuine and which exceptions count as merely apparent. However, that is consistent with acknowledging that other purported genre explanations, such as the postmodern dance one given earlier, we do have an adequate tacit understanding of the ceteris paribus clause. As emphasized earlier, specific genre explanations still need to be assessed on a case-by-case basis, through engagement with the relevant first-order debate. Even though the no-tacit-understanding thought may motivate the rejection of some specific genre explanations, it cannot motivate the robustness challenge to genre explanations as a kind.
4. The Informativeness Challenge
We now turn to the theoretical criticism that genre explanations, as a kind, are relatively uninformative. We can illustrate this criticism by returning to the two simple examples again. Why can't the appropriate response to Satisfyin’ Lover be explained by, say, Paxton's intentions in creating the choreography? Since Paxton's intention behind this piece is one factor that makes this piece a postmodern dance, it seems that an explanation that appeals to it must be as explanatory as, if not more explanatory than, an explanation that appeals to the genre of postmodern dance. Similarly, why can't the appropriate response to Evil Dead 2 be explained by, say, the specific ways in which it is similar to other specific horror comedies (e.g. the music cues used)? Since these specific resemblances are factors that make the movie a horror comedy, it seems that an explanation that appeals to the specific resemblances must be as explanatory as, if not more explanatory than, an explanation that appeals to the genre of horror comedy. Worryingly, genre explanations appear to be less informative than other aesthetic explanations – specifically those that appeal to the bases of genre classification.
The informativeness challenge directly questions the explanatory worth of genre: it alleges that genre explanations lack an essential feature of genuine explanations because they are relatively uninformative. It is driven by the smaller-grain intuition: explanations at lower levels give more information than explanations at higher levels.
All metaphysical accounts of genre acknowledge that a work's appropriate classification in a genre depends on some other features of the work, even if these accounts do not always agree on what the relevant features are. Call the potential bases for genre classification lower-level features. In contrast, genre is a higher-level feature because a work's appropriate classification in a genre depends on its lower-level features. Since genre is not metaphysically fundamental, it is tempting to think that it is not explanatorily fundamental either. Following Currie, we can contrast genre explanations with individualistic explanations, which denote in this context explanations that cite metaphysically fundamental features.Footnote 30 Individualistic explanations tell us more about the specifics of a work, especially how it differs from works that are similar in other respects, such as other works in the same genre. To answer the informativeness challenge, we must say what information genre explanations can provide that individualistic explanations cannot.
5. Genre Explanations and Informativeness
The smaller-grain intuition behind the informativeness challenge is misguided. Explanations at different levels provide different kinds of information, suitable for different interests and aims. Although physics is ontologically more fundamental than the social sciences, physical explanations are not always preferable because they can fail to provide information that higher-level social scientific explanations provide. Similarly, I argue, although genre classification depends on lower-level features, individualistic explanations need not be always preferable because they can fail to provide information that genre explanations provide. In this section, I present explanatory pluralism, the view that explanations at different levels can be valuable for different interests and aims, and then consider how it helps in responding to the informativeness challenge.
5.1. Explanatory Pluralism
Explanatory pluralism says that explanations at different levels are valuable for different interests and aims. On the version developed by Angela Potochnik, this is because explanations at different levels illuminate different patterns of dependence – typically in the sciences, patterns of causal relationships – that hold in different circumstances.Footnote 31
To illustrate, consider Potochnik's example:
Population genetics and evolutionary ecology provide competing explanations of why many traits evolve. Genetic [lower-level] explanations show how the distribution of genotypes related to the trait changed generation by generation, whereas phenotypic [higher-level] explanations show how the environment selectively advantaged the trait in question. […] The two explanations cite different properties, at different levels of organization, to explain the same event.Footnote 32
Genetic and phenotypic explanations illuminate different patterns of dependence. A genetic explanation of a trait's evolutionary development can withstand variances in the environmental source of fitness. A phenotypic explanation of a trait's evolutionary development can withstand variances in the genetic details. The two patterns of dependence do not perfectly overlap; they hold in different ranges of circumstances.
Patterns of dependence illuminated by lower- and higher-level explanations do not overlap when properties cited in higher-level explanations do not supervene on just the properties cited in lower-level explanation:
Yet phenotypes do not supervene on genes or genotypes but on a complex combination of properties including many other properties of the organism, properties of the environment, and sometimes even properties of other organisms. The lower-level properties under investigation – genes – are not proper candidates for the supervenience bases of phenotypes.Footnote 33
It is because phenotypes do not supervene on just genotypes that there can be variances in phenotypes that are due only to variances in its other supervenience bases, such as properties of the environment. Genetic explanations cannot capture these phenotypic variances. Moreover, there can also be variances in genotypes that are not reflected by variances in phenotypes, perhaps due to the constancy in the phenotypes' other supervenience bases. Phenotypic explanations cannot capture these genotypic variances.
Different interests and aims call for different patterns of dependence. If we want to know how organisms of different species in the same environment can share the same trait, then a phenotypic explanation is more informative. If we want to know how organisms of the same species in different environments can share the same trait, then a genetic explanation is more informative. Given that the patterns of dependence that these explanations illuminate hold in different circumstances, neither explanation is more general than the other. (The present point about generality thus echoes the point about counterfactual robustness in §3.) Therefore, neither explanation can always be said to be more informative than the other because each has a range of applicability different from the other's.
Explanatory pluralism's sensitivity to interests and aims does not deny that lower-level properties are more fundamental than higher-level properties in some sense. After all, phenotypes do partly depend on genes. What explanatory pluralism denies is that this kind of fundamentality implies, or is equivalent to, explanatory fundamentality. Since lower- and higher-level explanations can give us different information about distinctive patterns of dependence, neither kind of information ought to be uniformly preferred over the other, contrary to what the smaller-grain intuition suggests.Footnote 34
5.2. Response to the Informativeness Challenge
Now, let us return to the informativeness challenge. To answer this challenge, we must show that genre explanations can provide information that individualistic explanations cannot. Explanatory pluralism shows us how.
To develop a strategy for responding to the informativeness challenge, we start with the horror comedy case from §1. There are (at least) two aesthetic explanations that one can give for why laughter is the appropriate response to a decapitation scene in Evil Dead 2. First, one can give a genre (higher-level) explanation that appeals to the generalization that decapitation scenes in horror comedies warrant laughter. (We are supposing that decapitation scenes in horror comedies warrant laughter is a genre law.) Second, one can give an individualistic (lower-level) explanation that appeals to ‘the specific way [the work] is (together with, perhaps, similarities between these specific ways and specific ways possessed by other specific works the people in the audience are familiar with)’.Footnote 35
What is the relationship between the properties cited in the higher-level aesthetic explanation – genre classifications – and the properties cited in the lower-level aesthetic explanation – specific resemblances to other horror comedies? Although there are disagreements about the details, philosophers agree that genre classifications, like phenotypes, supervene on a complex combination of properties. For example, Kendall Walton says that a work's appropriate classification in a genre depends on its relevant resemblances to other works in that genre, authorial intention, critical judgment, and that genre's propensity for increasing aesthetic pleasure.Footnote 36 In other words, Evil Dead 2’s appropriate classification as a horror comedy is due to a complex combination of factors that includes its specific resemblances to other horror comedies. The relationship between the properties cited in the two aesthetic explanations above parallels the relationship between the properties cited in phenotypic and genetic explanations.
Since genre classifications do not supervene on just specific resemblances, there can be variances in genre classifications that are only due to variances in its other supervenience bases, such as authorial intention. An individualistic explanation that cites only specific resemblances cannot capture these genre classification variances. Therefore, in this case, the genre explanation illuminates a pattern of dependence that differs from the ones that the individualistic explanation illuminates.
Consider again why we should laugh at a decapitation scene in Evil Dead 2. The genre explanation holds in circumstances where the individualistic explanation does not. It tells us why we should also laugh at a decapitation scene in, say, Shaun of the Dead even though that movie does not contain the same specific resemblances to other horror comedies as Evil Dead 2. Furthermore, it tells us why we should still laugh at a decapitation scene in Evil Dead 2 even if some of the specific works that it is similar to never existed. These limitations of the individualistic explanation show that the genre explanation can be preferable for some interests and aims.
We now have a strategy for responding to the informativeness challenge: recognizing that genre explanations can illuminate patterns of dependence that individualistic explanations cannot. Importantly, we can acknowledge that genre classifications depend on lower-level properties, such as specific resemblances, while denying that explanations that cite lower-level properties are explanatorily fundamental. Explanatory pluralism situates genre explanations alongside other aesthetic explanations, including individualistic explanations. Aesthetic explanations at different levels are complementary in the same way that scientific explanations at different levels are. Rather than insisting that one kind of aesthetic explanation is uniformly superior to another, we should use whichever kind of aesthetic explanation best suits our aims and interests in a given context.
Our response to the informativeness challenge shows that whether a genre explanation is a good aesthetic explanation depends on whether it can provide information about a phenomenon that an individualistic explanation of the same phenomenon cannot. Genre explanations cannot be ruled out as good aesthetic explanations simply because they are higher-level explanations. So, the a priori and categorical rejection of genre explanations on dependence fundamentality grounds is not justified. To reiterate the earlier slogan: evaluating the worth of a specific genre explanation requires doing aesthetics, not meta-aesthetics.
6. Concluding Remarks
6.1. Specifics
With the resources developed in previous sections, we now return to Currie's arguments against genre explanations, presented in §1. Currie makes two implicit assumptions in his arguments. First, Currie assumes that it is unproblematic to switch the comparison class for genre explanations between the two objections. Genre explanations are initially compared to lower-level aesthetic explanations and then compared to higher-level sociological explanations. Second, Currie assumes that it makes sense to talk about a unique best kind of explanation irrespective of the questions that are asked. Although he only mentions one particular question one could ask, about why Hamlet is so intellectually and emotionally affecting, he draws the categorical conclusion that genre generally does not figure in ‘explanations that have any artistic or aesthetic interest’.Footnote 37 Both assumptions turn out to be problematic.
As §5 argues, there is a pragmatic dimension to explanations. Whether an explanation is suitable depends on contextual factors such as the question asked and the aims and interests of the questioner. In answering questions about aesthetic phenomena, sociological explanations are simply irrelevant. Hence, even if Currie were right that genre explanations are relatively less robust than higher-level sociological explanations, in answering questions about aesthetic phenomena, higher-order sociological explanations are simply not in the salient comparison class. Only comparisons to individualistic aesthetic explanations, and not comparisons to higher-level sociological explanations, are relevant for assessing the robustness of genre explanations.
However, as §3 argues, Currie is also wrong in assuming that genre explanations are relatively less robust than higher-level sociological explanations. Both kinds of explanations are counterfactually robust despite the exceptions that they apparently admit of. The range of counterfactuals that genre laws support is no narrower than the range of counterfactuals that sociological laws support. Specifically, there are scenarios where a genre law holds but a sociological law does not. Suppose that the psychology of human beings is the same as it actually is, except that human beings are incapable of following organized religions. Then, in this scenario, it would still be true that decapitation scenes in horror comedies warrant laughter but false that the popularity of organized religions declines when the population shifts due to industrialization. Given that each discipline is autonomous, each discipline's laws support counterfactuals that range over a distinct set of scenarios.
6.2. Generalities
Both challenges to genre explanations come from misguided armchair intuitions. We can correct these armchair intuitions by paying attention to central insights from recent works on scientific explanations. The no-exception intuition behind the robustness challenge loses its force once we recognize the centrality of ceteris paribus laws in actual scientific practice. The smaller-grain intuition behind the informativeness challenge loses its force once we recognize the need for explanatory pluralism in the special sciences. These insights also allow us to transform the specific responses to Currie's arguments into general lessons for better understanding aesthetic explanations.
First, we should be pragmatists. What allows for a satisfying response to a given question depends on contextual factors, such as the aims and interests of a research program. There is a pragmatic dimension to which explanation counts as best. In assessing the worth of an explanation, what matters is how it measures up to other competing explanations in the same research program.
Second, we should be pluralists. Aesthetic explanations at different levels, like scientific explanations at different levels, illuminate different patterns of dependence. The pragmatic dimension of explanations suggests that explanations at different levels are good for answering questions with different aims and interests. We should not uniformly prefer explanations at one level to explanations at another.
